Neuronal and synaptic mechanisms of associative memory

Brains can hold a remarkable amount of information within their networks of interconnected neurons. We often think of associative memory as a graph, with nodes representing objects or events, and edges (lines) representing the associations between them. However, little is known of how complex representations, such as categories or place-reward associations, are learned and encoded by the activity and connectivity of neuronal assemblies.

We address these questions at the cellular and circuit level, by performing long-term calcium imaging experiments in the primary and higher visual cortical areas and in the medial prefrontal cortex of the mouse. We make use of two-photon microscopy to image calcium activity of large groups of neurons throughout the process of learning visual categories. By quantifying changes in function of neurons within specific brain areas, we aim to determine different stages of the process of forming abstract/complex representations.

Conventional two-photon microscopes are large and heavy setups and cannot easily be adopted to behavioral paradigms that require the mouse to roam-around freely, such as place-reward learning. Therefore, we additionally make use of a head-mounted miniaturized epifluorescence microscope that the mouse can carry around while performing maze-learning tasks. 

Using these approaches, we identify specific cells and circuits in the brain that are involved in learning of episodic or semantic associations. In subsequent steps, we aim to trace connectivity within, and the inputs to, these circuits in order to better understand how learned representations arise from changes in the underlying synaptic connectivity.

Pieter Goltstein


Sandra Reinert

PhD Student

Joanna Klos

PhD Student

Mark Hübener

Research Group Leader

Max Planck Institute for biological Intelligence

Am Klopferspitz 18
82152 Munich-Martinsried